Remote Sensing Images In Building Recognition And Change Detection

The artificial surface features are the important element of the special geographical information shed. The artificial surface features mainly include building, bridge, road and large-scale engineering building .And 80 percents of high distinguish remote sensing image of metropolitan district is building and road. Server as the main substance of surface features category, the identify and withdraw of building occupies very big proportion. So, how to identify and withdraw the building of remote sensing image is one of the most important research subjects of target identify of remote sensing image. The work including:First, begin with the roof of building, mainly introduces how to born the roof of building . In my brochure, some existent algorithms have been recruited and improved, and finally proved by experiments.Second, because of the require of military application, not only we should withdraw and identify the building accurately, but also analyses the variation occasion of the building witch have been stricken. So, we should research the variation occasion at different time. There are two methods of identifying injure of the building:1. Based on image matching. This method mainly adopts related modulus.2. Based on texture. This method uses texture decrease to evaluate the variation of the building .And it was lack of precision of location and cost too much computation time.Using image matching based on relativity, this method eliminates some effect of illumination and climate, and abates the fault probabilistic of identify, increases the precision of direct current position. The result proved this algorithm is better than the one based on texture.